Can you OCR a WEBP image on Windows?
Yes. Read by Zaweba reads text from WEBP images through Windows' own codecs, and if a Store extension is needed it tells you which one to add.
Yes, you can. Read by Zaweba reads the text in WEBP images on Windows, using Windows' own image codecs. Some PCs need the WEBP extension from the Microsoft Store before Windows can decode the format, and if yours does, Read tells you so plainly instead of throwing an error you have to puzzle out.
WEBP is the format a lot of websites now serve their pictures in. Save an infographic, a product image or a slide from a web page and there's a good chance it lands as a .webp file. That's fine until you want the words in it.
How Read opens a WEBP file
Read by Zaweba doesn't bundle its own decoder for every picture format. It asks Windows to open the file, the same way the Photos app does. PNG, JPEG, TIFF, BMP and GIF are built in to Windows. WEBP and HEIC rely on extensions from the Store. When Windows has the extension, Read opens the WEBP like any other picture. When it doesn't, Read names what's missing so you can add it and try again.
Once open, a WEBP is treated exactly like a PNG or a JPEG. It goes through the same clean-up, the same choice of engine and the same review.
Working through it
Save the WEBP image to your PC, then drop it onto Read's window. If you've copied the picture from a browser, you can often just press Ctrl+V in Read to paste it from the clipboard, which skips saving altogether.
For a single picture like this, Read's Auto setting usually chooses Windows OCR when Windows has the language, because it's quick. If the image has columns or a table, Auto picks Tesseract instead, as it does layout analysis. The engine used and the reason are shown above the text, and Read again with Tesseract is on hand if you want a second opinion with word confidences.
Check the text against the picture in the side-by-side view, make any corrections, then copy it or save it. Plain text, .docx, searchable PDF and CSV for tables are all on offer.
What to expect from web images
Web pictures are often small. A heading in a banner might be plenty big, while the caption underneath is a few pixels high. Read's Help tab is upfront that low resolution pictures and decorative fonts are unreliable, and fancy typography is common on the web. If the reading is patchy, look for a larger version of the image, or zoom the page in the browser and grab the text from the screen instead.
That screen grab is often the easier route anyway. Press Read's hotkey (Win+Shift+T by default), drag round the text, and it's copied. Read enlarges the area to twice its size before reading because screen text is small, which helps with exactly this kind of picture.
Keeping it on your PC
Online converters will happily take a WEBP and hand back text, but you're uploading the picture to do it. Read by Zaweba does all of the reading on your own machine. The only thing it ever downloads is a language pack, and only when you click to get one. For a screenshot of a private message or a company slide, that's worth having.
